Microsoft alone could solve the world's hunger problems. They could cure all the curable diseases. They could save the 33,000 kids that are dying every day because they're hungry.
Huh? Are you stupid, or just dilusional? That was the most ridiculous thing I've read in quite some time.
Besides, Gates and Microsoft have donated plenty. Their donations -- both as a percentage of profit and in absolute terms -- dwarf those of most other companies, including those that are many times larger than Microsoft, such as Walmart.
Seems to me like they're going down the list of things people have been asking for. After a five year hiatus I'm just glad the IE team is back in business. Competition is good for everybody, even if you're a FireFox or Safari user.
The problem with calling them "undocumented" or "secret" APIs is that those are loaded terms, implying that Microsoft is somehow using them internally in a way that gives them a competitive advantage. We can quibble over naming, but what I am saying is that it is common (practically unavoidable, actually) for an OS (and applications) to expose functionality -- as DLL exports, COM objects, and the like -- that is consumed by other internal OS components or that same application. There's absolutely nothing wrong with this technically, legally, or morally.
Documenting these is actually a bad thing, as they are internal implementation details that fluctuate from release to release. Additionally, they are not factored or tested for public consumption, and documenting them would be very noisy and chaotic for anybody that wants to do development on the platform. They also create huge application compatibility headaches, no matter who uses them or how they are discovered.
The crux of the alleged "secret API" issue is whether Microsoft is allowing its application developers to use this functionality while withholding it from competitors. You and I both agree this is wrong, and Microsoft should be called on it if and when it happens. The fact that all of these exports, such as the referenced article are in such plain view makes it fairly easy to trace if a competitor suspects something.
By the way, the burden of proof is not on me to show that each one of those hundreds of thousands of entry points are NOT called by any application that Microsoft has ever created. The burden of proof is on those making the claim to prove that Microsoft is unfairly using them.
Also, if you're feeling overconfident, go to a Ars Technica's BattleFront forum and start a "secret API" topic.:-)
So I know that the distinction you're making between an "API" and "mere" public functions is false - if a class/object member is exposed for calling/reading/writing, it is an Interface for Programming the Application. And if MS hasn't published descriptions of its function and arguments, it's undocumented.
I guess we will have to agree to disagree. There's nothing more I can say to convince you, and as someone who has worked as a developer on Windows it is my assertion that your statement is absolutely false.
Now go pull up Dependency Walker and OleView and see the hundreds of thousands of exported functions, COM objects, and interfaces that are part of the internal workings of the OS, not documented, but nonetheless "exposed". You can try to call them on it, but you're not going to get very far because your technical understandings of the facts are incorrect.
(Actually, just "link/dump/exports c:\windows\system32\*.dll" should keep you busy for a long while.)
The fundamental thing that you're failing to grasp is that exporting a function from a DLL is not creating an "undocumented API".
Code can be also be shared by creating a library, and statically linking. Code can also be shared by creating a COM object. This *is* encapsulation and componentization, but just one level. It's arbitrary -- and completely incorrect -- to say that once you share code across files it becomes an "undocumented API".
Since it seems that you've never done any large scale programming, let me put this another way. API means Application Programming Interface. If an exported function (or COM object, or whatever) is not consumed by another app, it's not an API.
I'm in complete agreement with the philosophy that Microsoft applications should only use documented APIs. (And so should Apple's, etc.)
Let me try to be even more clear. For example, if Office is calling a function exported from a DLL that ships with Windows and that function is not documented in MSDN, that is wrong. However, if WordUI.dll uses an exported function in WordCoreLogic.dll, there's nothing wrong with that.
It is trivially easy to set breakpoints and see if other apps or OS components are calling these APIs.
Exporting functions to share code is also known as "sharing code" not just "really bad programming" and "unstable hack", although based on some of the functions I've seen exported, "internet time" leads to the latter sometimes. Still no reason to document arbitrary useless internal helper functions.
Go use Dependency Walker and you will see hundreds -- nay, thousands! -- of exports from DLLs in system32 that aren't documented. They aren't documented because they aren't consumed by "middleware" -- they are exported and consumed by the OS, usually the same app/component that exported them. IE is now legally classified as middleware, I think, so it cannot use any APIs that aren't documented.
Undocumented APIs are APIs used by other applications. He just pulled out every exported function that is exposed by IE, even if it's just consumed by IE.
These are like a private function of a class, but shared across DLLs for logistical reasons.
As others point out, the user must be running as admin. All NT-based OSes have a very complete and robust security model.
I absolutely agree that Microsoft needs to be much more hardcore in stopping people running as Admin. The only reason they don't is too many apps break; a chicken-and-egg problem that Microsoft is, ironically, being too timid to stop.
I'm certainly not a lawyer, but I think the overall point is that EULAs are written in that way because companies do not want to make a guarantee that their software is fit for any given (or specific) purpose.
This says nothing about the actual or attempted backing: if I release a piece of freeware I wrote in my spare time, and stick the same EULA on it that Windows has, Windows still has better support. End users can trust that Microsoft is going to going to actively support that OS through it's planned end of life, but I may never answer a single email or release a single update or patch for my freeware app.
In other words, EULAs are simply a necessary evil in our overly-litigious society. You need to look at companies' actual track record to predict the level of support you're likely to get. Depending on your situation and personal opinion you will have a different opinion whether OSS-style or MS-style support is better, but it's all about the track record, not the EULA.
Lotito's performances are the consumption of metal, glass, rubber and so on in items such as bicycles, televisions, a Cessna 150, and smaller items which are disassembled, cut-up and swallowed. The aircraft took roughly two years to be 'eaten' from 1978 to 1980. He began eating unusual material while a child and has been performing publicly since 1966.
Disks are a huge bottleneck, especially when the OS has to do a hard page fault. Therefore, at least in the client world, we still spend an enormous amount of time optimizing code to minimize working set and IO.
This illustrates a common fallacy...Effort spent does not automatically qualify you for a reward
Fine, replace "know" with "speculate" in my first part. The point is, if you would be less pedantic for a second, that IP enables the potential for reward. That potential encourages more people to take risks. And that, like most modern nontrivial ventures, is why most software gets built.
If it's the compensation that makes it worth your while, then you're doing it for the money, which makes it unlikely that you'll produce the best possible Killer App for whatever task it is meant to perform.
Wow, as a logical thinker you sure pulled a boner there. Software is often created for love and money. The money also helps, you know, supply time to actually build the software without starving to death. (Didn't RMS one of those guys used to suggest software developers should develop at night, and be waiters and waitresses during the day? Oh, fun.)
Writing software professionally is like fishing, and if the fish figure out how to get the bait without biting the hook, that's just your bad luck.
Just like, if somebody figures out how to hack a bank and steal $400,000, it's that bank's bad luck? See my other reply; money essentially boils down to time (and work/effort). Oh, and if your attitude guided laws, machines would be locked down so fast it would make your head spin. Bring on the DRM, encryption, and welded tamper-resistant cases!
And since the argument has no solution so simple as to be uiniversally acceptable, it'll go on for a long time...I just see the current implementation of the ideals and ideas of intellectual property as seriously flawed, and would like them changed, rather than eliminated completely.
There's room for improvement, which is what actionable debate focuses on. However, for those that wish to essentially abolish copyright and patent laws, there is no real debate. Those people are on the lunatic fringe, which is why I can sleep at night instead of honing my french frying, buffalo spearing, corn growing, and real-estate selling skills.
Nice analogy, but it does not fit: If someone "steals" my intellectual property, none of my hard work or unique decisions are gone.
What you have failed to grasp, is that what is being stolen is time. In my analogy (all analogies are flawed, of course), you're free to start from scratch and get your degree(s)again.
You still have the full use of your creation.
Again, you're still thinking like a hunter-gatherer. Developing software, creating music, or filming a movie isn't possible if nobody pays: the economy of scale isn't there. Yes, there are people who write software and give it away for free -- I'm certainly not saying there's anything wrong with that! It's their creation; they get to decide how, and for what cost, their creation is distributed.
All very true, however intellectual "property" is different in that it has no physical form. This is significant. It means that I can make a copy of your intellectual "property" and not deprive you of the use of it.
So if I spend ten years of my life, 80 hours a week, dedicated to building The Killer App of the 21st Century, knowing that millions of people will want what I create and voluntarily give me compensation that makes it worth my while...
And then all those millions of people just steal* my creation from me (they want it; they can afford it; they have just learned they don't have to pay), I haven't been deprived of anything**? What planet are you from?
Why is hammering a nail worth getting paid for, but building complex systems that entertain people or save them millions of dollars not worth anything? Oh sure, you're going to say that supporting those complex systems are where we should make our money. What a stupid idea. Not only is it stupid because it's an indirect, roundabout way of making money, it's also stupid because you've still eliminated the incentive to create; the support companies will compete to create as little as possible, because that would incur a cost and put them at a disadvantage over time. It's also stupid because the last thing a customer wants is a product that is so difficult to use that it has to be supported, but that's what they're going to end up with.
Yeah, it's been said a billion times before. Although I see you're a researcher, so let me give you an analogy. Your stealing my years of hard work is roughly analogous to you getting a precision lobotomy (magical, if you will) that wipes out everything you have learned and worked for academically since high school. Your hard work? Good decisions? Gone. You're now like everyone else.
* Oh, I'm sorry, I meant to say "violate my copyright."
** I value my time at about $100/hr, so said app would have a personal opportunity cost to me of over $400,000. Even if I enjoyed creating it, it is my decision to give it away or not. That's what you have stolen, although modern software projects require far more than 10-20 man years of development cost.
P.S. I'm actually not bitter, as the above rant might suggest, because in the real world -- at least in developed countries -- intellectual property is still valued, despite the futile wishes of those that want it to go away.
Hahaha... A classic "Bill Parish" link. From 1999. Welcome to 2005.
Only now, Microsoft is expensing stock options and being as transparent as possible in its accounting, so the link simply doesn't hold water. Not that it ever did.
That is only one more reason not to work for Microsoft. Do they expect their employees to laugh? Like the post just after this one, I felt sorry for him. It's embarassing.
rbanffy, a while back you made the following statement. Have your forgotten?
One should never take himself or the world too seriously. Fundamentalists do it and we can all see what the consequences are.
Spoof videos are a tradition at Microsoft, both company-wide and within individual teams. Some are better than others, and they're all somewhat cheesy, but they're usually a good icebreaker and it's refreshing to see executives, management, and fellow employees willing to poke a little bit of fun at themselves. Quite often they involve inside jokes, and a better understanding of the personalities of the people in the video, so I guess it's not that surprising when they occasionally leak and people outside Microsoft don't get the joke.
Regardless, it helps keep employees from taking themselves too seriously.
Judging by a lot of/. comments, people here would be well advised to take your original advice. The religious ferocity that they defend open source and bash Microsoft and other entities that don't adhere to their world view -- even over silly things like spoof videos -- is rather... unhealthy.
Why offer this app when they can fix Windows to make it either very difficult or impossible for spyware/viruses to install. They're lazy.
Explain to me how we can prevent individual users from running software that may be "bad" without significantly degrading the user experience (including backwards compatibility), and I will make sure it's implemented...
A good point, especially when it comes to those lofty, "influential", graduation-style speeches.
But, as somebody who came from a family where nobody in the previous generation had gone to college, I would have appreciated more useful advice about college (and related preparation).
For example, I've done extremely well academically and career-wise, but I was overly-stressed and worried throughout high school and college, and didn't spend enough time just living and having fun. This led to under-socialization, which is probably the biggest problem in my life today.
Whereas my parents exacerbated this because they wanted me to be successful, peers who had parents (or other adults in their lives) with college experience had an advantage: they had a better understanding of how that world worked, what was really important, and what was coming.
The internet now makes it easy for people of all backgrounds to learn from the life experiences of others, although the noise probably drowns out a lot of the useful stuff.
Can any video experts comment how transcoding from MPEG-2 affects video quality?
This is a good test for comparing the quality of codecs for ripping DVDs, but do the results hold true when an uncompressed master is used as the source?
1) Single outer window with a single menubar and docking toolbars. (No, Deweirdifyer is not good enough.)
2) Use native dialogs where possible (GIMP's File/Open dialogs are the worst file dialogs I've seen in years, but the bigger issue is the inconsistency).
Both of these are trivial and constantly requested. If the GIMP developers do not implement them in the next release it will prove they're not interested in listening to their users.
Slashdot Mirror
Sadly it isn't paranoid... Remember the user-agent: opera stylesheet problem?
Give me a break. MSN made a mistake and Opera milked it for all it was worth.
MS is walking on eggshells; the likelihood of them doing anything malicious toward their competitors at this point is nil.
Microsoft alone could solve the world's hunger problems. They could cure all the curable diseases. They could save the 33,000 kids that are dying every day because they're hungry.
Huh? Are you stupid, or just dilusional? That was the most ridiculous thing I've read in quite some time.
Besides, Gates and Microsoft have donated plenty. Their donations -- both as a percentage of profit and in absolute terms -- dwarf those of most other companies, including those that are many times larger than Microsoft, such as Walmart.
Perhaps you missed it, but they have announced support for PNG alpha transparency: http://blogs.msdn.com/ie/archive/2005/04/26/412263 .aspx
Seems to me like they're going down the list of things people have been asking for. After a five year hiatus I'm just glad the IE team is back in business. Competition is good for everybody, even if you're a FireFox or Safari user.
What's stoping you from writing it yourself? It's probably only a couple hundred man years of effort; certainly worth doing to save $500, right?"
The problem with calling them "undocumented" or "secret" APIs is that those are loaded terms, implying that Microsoft is somehow using them internally in a way that gives them a competitive advantage. We can quibble over naming, but what I am saying is that it is common (practically unavoidable, actually) for an OS (and applications) to expose functionality -- as DLL exports, COM objects, and the like -- that is consumed by other internal OS components or that same application. There's absolutely nothing wrong with this technically, legally, or morally.
:-)
Documenting these is actually a bad thing, as they are internal implementation details that fluctuate from release to release. Additionally, they are not factored or tested for public consumption, and documenting them would be very noisy and chaotic for anybody that wants to do development on the platform. They also create huge application compatibility headaches, no matter who uses them or how they are discovered.
The crux of the alleged "secret API" issue is whether Microsoft is allowing its application developers to use this functionality while withholding it from competitors. You and I both agree this is wrong, and Microsoft should be called on it if and when it happens. The fact that all of these exports, such as the referenced article are in such plain view makes it fairly easy to trace if a competitor suspects something.
By the way, the burden of proof is not on me to show that each one of those hundreds of thousands of entry points are NOT called by any application that Microsoft has ever created. The burden of proof is on those making the claim to prove that Microsoft is unfairly using them.
Also, if you're feeling overconfident, go to a Ars Technica's BattleFront forum and start a "secret API" topic.
So I know that the distinction you're making between an "API" and "mere" public functions is false - if a class/object member is exposed for calling/reading/writing, it is an Interface for Programming the Application. And if MS hasn't published descriptions of its function and arguments, it's undocumented.
I guess we will have to agree to disagree. There's nothing more I can say to convince you, and as someone who has worked as a developer on Windows it is my assertion that your statement is absolutely false.
Now go pull up Dependency Walker and OleView and see the hundreds of thousands of exported functions, COM objects, and interfaces that are part of the internal workings of the OS, not documented, but nonetheless "exposed". You can try to call them on it, but you're not going to get very far because your technical understandings of the facts are incorrect.
(Actually, just "link /dump /exports c:\windows\system32\*.dll" should keep you busy for a long while.)
The fundamental thing that you're failing to grasp is that exporting a function from a DLL is not creating an "undocumented API".
Code can be also be shared by creating a library, and statically linking. Code can also be shared by creating a COM object. This *is* encapsulation and componentization, but just one level. It's arbitrary -- and completely incorrect -- to say that once you share code across files it becomes an "undocumented API".
Since it seems that you've never done any large scale programming, let me put this another way. API means Application Programming Interface. If an exported function (or COM object, or whatever) is not consumed by another app, it's not an API.
I'm in complete agreement with the philosophy that Microsoft applications should only use documented APIs. (And so should Apple's, etc.)
Let me try to be even more clear. For example, if Office is calling a function exported from a DLL that ships with Windows and that function is not documented in MSDN, that is wrong. However, if WordUI.dll uses an exported function in WordCoreLogic.dll, there's nothing wrong with that.
It is trivially easy to set breakpoints and see if other apps or OS components are calling these APIs.
Exporting functions to share code is also known as "sharing code" not just "really bad programming" and "unstable hack", although based on some of the functions I've seen exported, "internet time" leads to the latter sometimes. Still no reason to document arbitrary useless internal helper functions.
Go use Dependency Walker and you will see hundreds -- nay, thousands! -- of exports from DLLs in system32 that aren't documented. They aren't documented because they aren't consumed by "middleware" -- they are exported and consumed by the OS, usually the same app/component that exported them. IE is now legally classified as middleware, I think, so it cannot use any APIs that aren't documented.
And the other three /. employees are super-bitter because they were fired from Microsoft. ;-)
Because it's wrong.
Undocumented APIs are APIs used by other applications. He just pulled out every exported function that is exposed by IE, even if it's just consumed by IE.
These are like a private function of a class, but shared across DLLs for logistical reasons.
As others point out, the user must be running as admin. All NT-based OSes have a very complete and robust security model.
I absolutely agree that Microsoft needs to be much more hardcore in stopping people running as Admin. The only reason they don't is too many apps break; a chicken-and-egg problem that Microsoft is, ironically, being too timid to stop.
I'm certainly not a lawyer, but I think the overall point is that EULAs are written in that way because companies do not want to make a guarantee that their software is fit for any given (or specific) purpose.
This says nothing about the actual or attempted backing: if I release a piece of freeware I wrote in my spare time, and stick the same EULA on it that Windows has, Windows still has better support. End users can trust that Microsoft is going to going to actively support that OS through it's planned end of life, but I may never answer a single email or release a single update or patch for my freeware app.
In other words, EULAs are simply a necessary evil in our overly-litigious society. You need to look at companies' actual track record to predict the level of support you're likely to get. Depending on your situation and personal opinion you will have a different opinion whether OSS-style or MS-style support is better, but it's all about the track record, not the EULA.
Zen, your Google-fu is weak: http://en.wikipedia.org/wiki/Michel_Lotito :)
Agreed. Google's UI and rendering technology is far better than map24. A very impressive web app.
However, in general I'd still rather have a client-side solution such as Streets and Trips.
True.
False.
True.
False.
Disks are a huge bottleneck, especially when the OS has to do a hard page fault. Therefore, at least in the client world, we still spend an enormous amount of time optimizing code to minimize working set and IO.
This illustrates a common fallacy...Effort spent does not automatically qualify you for a reward
Fine, replace "know" with "speculate" in my first part. The point is, if you would be less pedantic for a second, that IP enables the potential for reward. That potential encourages more people to take risks. And that, like most modern nontrivial ventures, is why most software gets built.
If it's the compensation that makes it worth your while, then you're doing it for the money, which makes it unlikely that you'll produce the best possible Killer App for whatever task it is meant to perform.
Wow, as a logical thinker you sure pulled a boner there. Software is often created for love and money. The money also helps, you know, supply time to actually build the software without starving to death. (Didn't RMS one of those guys used to suggest software developers should develop at night, and be waiters and waitresses during the day? Oh, fun.)
Writing software professionally is like fishing, and if the fish figure out how to get the bait without biting the hook, that's just your bad luck.
Just like, if somebody figures out how to hack a bank and steal $400,000, it's that bank's bad luck? See my other reply; money essentially boils down to time (and work/effort). Oh, and if your attitude guided laws, machines would be locked down so fast it would make your head spin. Bring on the DRM, encryption, and welded tamper-resistant cases!
And since the argument has no solution so simple as to be uiniversally acceptable, it'll go on for a long time...I just see the current implementation of the ideals and ideas of intellectual property as seriously flawed, and would like them changed, rather than eliminated completely.
There's room for improvement, which is what actionable debate focuses on. However, for those that wish to essentially abolish copyright and patent laws, there is no real debate. Those people are on the lunatic fringe, which is why I can sleep at night instead of honing my french frying, buffalo spearing, corn growing, and real-estate selling skills.
Nice analogy, but it does not fit: If someone "steals" my intellectual property, none of my hard work or unique decisions are gone.
What you have failed to grasp, is that what is being stolen is time. In my analogy (all analogies are flawed, of course), you're free to start from scratch and get your degree(s)again.
You still have the full use of your creation.
Again, you're still thinking like a hunter-gatherer. Developing software, creating music, or filming a movie isn't possible if nobody pays: the economy of scale isn't there. Yes, there are people who write software and give it away for free -- I'm certainly not saying there's anything wrong with that! It's their creation; they get to decide how, and for what cost, their creation is distributed.
All very true, however intellectual "property" is different in that it has no physical form. This is significant. It means that I can make a copy of your intellectual "property" and not deprive you of the use of it.
So if I spend ten years of my life, 80 hours a week, dedicated to building The Killer App of the 21st Century, knowing that millions of people will want what I create and voluntarily give me compensation that makes it worth my while...
And then all those millions of people just steal* my creation from me (they want it; they can afford it; they have just learned they don't have to pay), I haven't been deprived of anything**? What planet are you from?
Why is hammering a nail worth getting paid for, but building complex systems that entertain people or save them millions of dollars not worth anything? Oh sure, you're going to say that supporting those complex systems are where we should make our money. What a stupid idea. Not only is it stupid because it's an indirect, roundabout way of making money, it's also stupid because you've still eliminated the incentive to create; the support companies will compete to create as little as possible, because that would incur a cost and put them at a disadvantage over time. It's also stupid because the last thing a customer wants is a product that is so difficult to use that it has to be supported, but that's what they're going to end up with.
Yeah, it's been said a billion times before. Although I see you're a researcher, so let me give you an analogy. Your stealing my years of hard work is roughly analogous to you getting a precision lobotomy (magical, if you will) that wipes out everything you have learned and worked for academically since high school. Your hard work? Good decisions? Gone. You're now like everyone else.
* Oh, I'm sorry, I meant to say "violate my copyright."
** I value my time at about $100/hr, so said app would have a personal opportunity cost to me of over $400,000. Even if I enjoyed creating it, it is my decision to give it away or not. That's what you have stolen, although modern software projects require far more than 10-20 man years of development cost.
P.S. I'm actually not bitter, as the above rant might suggest, because in the real world -- at least in developed countries -- intellectual property is still valued, despite the futile wishes of those that want it to go away.
Hahaha... A classic "Bill Parish" link. From 1999. Welcome to 2005.
Only now, Microsoft is expensing stock options and being as transparent as possible in its accounting, so the link simply doesn't hold water. Not that it ever did.
That is only one more reason not to work for Microsoft. Do they expect their employees to laugh? Like the post just after this one, I felt sorry for him. It's embarassing.
rbanffy, a while back you made the following statement. Have your forgotten?
Spoof videos are a tradition at Microsoft, both company-wide and within individual teams. Some are better than others, and they're all somewhat cheesy, but they're usually a good icebreaker and it's refreshing to see executives, management, and fellow employees willing to poke a little bit of fun at themselves. Quite often they involve inside jokes, and a better understanding of the personalities of the people in the video, so I guess it's not that surprising when they occasionally leak and people outside Microsoft don't get the joke.
Regardless, it helps keep employees from taking themselves too seriously.
Judging by a lot of /. comments, people here would be well advised to take your original advice. The religious ferocity that they defend open source and bash Microsoft and other entities that don't adhere to their world view -- even over silly things like spoof videos -- is rather... unhealthy.
Why offer this app when they can fix Windows to make it either very difficult or impossible for spyware/viruses to install. They're lazy.
Explain to me how we can prevent individual users from running software that may be "bad" without significantly degrading the user experience (including backwards compatibility), and I will make sure it's implemented...
A good point, especially when it comes to those lofty, "influential", graduation-style speeches.
But, as somebody who came from a family where nobody in the previous generation had gone to college, I would have appreciated more useful advice about college (and related preparation).
For example, I've done extremely well academically and career-wise, but I was overly-stressed and worried throughout high school and college, and didn't spend enough time just living and having fun. This led to under-socialization, which is probably the biggest problem in my life today.
Whereas my parents exacerbated this because they wanted me to be successful, peers who had parents (or other adults in their lives) with college experience had an advantage: they had a better understanding of how that world worked, what was really important, and what was coming.
The internet now makes it easy for people of all backgrounds to learn from the life experiences of others, although the noise probably drowns out a lot of the useful stuff.
I don't think they are copying either windows or mac, they are merely following what have been the unix way for the last 10 years.
Which is why Unix has been such a resounding success on the desktop for the last ten years, right?
Can any video experts comment how transcoding from MPEG-2 affects video quality?
This is a good test for comparing the quality of codecs for ripping DVDs, but do the results hold true when an uncompressed master is used as the source?
1) Single outer window with a single menubar and docking toolbars. (No, Deweirdifyer is not good enough.)
2) Use native dialogs where possible (GIMP's File/Open dialogs are the worst file dialogs I've seen in years, but the bigger issue is the inconsistency).
Both of these are trivial and constantly requested. If the GIMP developers do not implement them in the next release it will prove they're not interested in listening to their users.